2 # = lib/evo/io/parser/hmmscan_domain_extractor.rb - HmmscanDomainExtractor class
4 # Copyright:: Copyright (C) 2012 Christian M. Zmasek
5 # License:: GNU Lesser General Public License (LGPL)
10 require 'lib/evo/util/constants'
11 require 'lib/evo/msa/msa_factory'
12 require 'lib/evo/io/msa_io'
13 require 'lib/evo/io/writer/fasta_writer'
14 require 'lib/evo/io/parser/fasta_parser'
19 class HmmscanDomainExtractor
26 # raises ArgumentError, IOError, StandardError
37 add_domain_number_as_digit,
38 add_domain_number_as_letter,
44 Util.check_file_for_readability( hmmsearch_output )
45 Util.check_file_for_readability( fasta_sequence_file )
46 Util.check_file_for_writability( outfile )
47 Util.check_file_for_writability( passed_seqs_outfile )
48 Util.check_file_for_writability( failed_seqs_outfile )
51 factory = MsaFactory.new()
52 in_msa = factory.create_msa_from_file( fasta_sequence_file, FastaParser.new() )
54 if ( in_msa == nil || in_msa.get_number_of_seqs() < 1 )
55 error_msg = "could not find fasta sequences in " + fasta_sequence_file
56 raise IOError, error_msg
64 out_msa_distant_partners = nil
65 out_msa_singlets = nil
67 out_msa_pairs = Msa.new
68 out_msa_distant_partners = Msa.new
69 out_msa_singlets = Msa.new
72 ld = Constants::LINE_DELIMITER
74 domain_pass_counter = 0
75 domain_fail_counter = 0
77 distant_pairs_counter = 0
78 close_pairs_counter = 0
79 proteins_with_passing_domains = 0
80 proteins_with_failing_domains = 0
81 max_domain_copy_number_per_protein = -1
82 max_domain_copy_number_sequence = ""
88 # prev_i_e_value = nil
91 File.open( hmmsearch_output ) do | file |
92 while line = file.gets
93 if !is_ignorable?( line ) && line =~ /^\S+\s+/
95 # tn acc tlen query acc qlen Evalue score bias # of c-E i-E score bias hf ht af at ef et acc desc
96 # 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
97 line =~ /^(\S+)\s+(\S+)\s+(\d+)\s+(\S+)\s+(\S+)\s+(\d+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\d+)\s+(\d+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\S+)\s+(\d+)\s+(\d+)\s+(\d+)\s+(\d+)\s+(\d+)\s+(\d+)\s+(\S+)\s+(.*)/
100 if domain_id != target_name
110 if ( number > max_domain_copy_number_per_protein )
111 max_domain_copy_number_sequence = sequence
112 max_domain_copy_number_per_protein = number
114 if ( ( ( e_value_threshold < 0.0 ) || ( i_e_value <= e_value_threshold ) ) &&
115 ( ( length_threshold <= 0 ) || ( env_to - env_from + 1 ) >= length_threshold.to_f ) )
117 extract_domain( sequence,
126 add_domain_number_as_digit,
127 add_domain_number_as_letter,
130 domain_pass_counter += 1
132 if passed_seqs.find_by_name_start( sequence, true ).length < 1
133 add_sequence( sequence, in_msa, passed_seqs )
134 proteins_with_passing_domains += 1
139 extract_domain( sequence,
152 singlets_counter += 1
154 if sequence != prev_sequence
157 if ( env_from - prev_env_to ) <= min_linker
158 extract_domain( sequence,
159 prev_number.to_s + "+" + number.to_s,
172 close_pairs_counter += 2
175 extract_domain( sequence,
181 out_msa_distant_partners,
188 distant_pairs_counter += 1
191 extract_domain( sequence,
197 out_msa_distant_partners,
204 distant_pairs_counter += 1
208 end # sequence != prev_sequence else
210 prev_sequence = sequence
212 prev_env_from = env_from
214 #prev_i_e_value = i_e_value
218 print( domain_fail_counter.to_s + ": " + sequence.to_s + " did not meet threshold(s)" )
219 log << domain_fail_counter.to_s + ": " + sequence.to_s + " did not meet threshold(s)"
220 if ( ( e_value_threshold.to_f >= 0.0 ) && ( i_e_value > e_value_threshold ) )
221 print( " iE=" + i_e_value.to_s )
222 log << " iE=" + i_e_value.to_s
224 if ( ( length_threshold.to_f > 0 ) && ( env_to - env_from + 1 ) < length_threshold.to_f )
225 le = env_to - env_from + 1
226 print( " l=" + le.to_s )
227 log << " l=" + le.to_s
229 print( Constants::LINE_DELIMITER )
230 log << Constants::LINE_DELIMITER
231 domain_fail_counter += 1
233 if failed_seqs.find_by_name_start( sequence, true ).length < 1
234 add_sequence( sequence, in_msa, failed_seqs )
235 proteins_with_failing_domains += 1
239 end # if !is_ignorable?( line ) && line =~ /^\S+\s+/
240 end # while line = file.gets
241 end # File.open( hmmsearch_output ) do | file |
243 if domain_pass_counter < 1
244 error_msg = "no domain sequences were extracted"
245 raise StandardError, error_msg
248 log << Constants::LINE_DELIMITER
249 puts( "Max domain copy number per protein : " + max_domain_copy_number_per_protein.to_s )
250 log << "Max domain copy number per protein : " + max_domain_copy_number_per_protein.to_s
251 log << Constants::LINE_DELIMITER
253 if ( max_domain_copy_number_per_protein > 1 )
254 puts( "First protein with this copy number: " + max_domain_copy_number_sequence )
255 log << "First protein with this copy number: " + max_domain_copy_number_sequence
256 log << Constants::LINE_DELIMITER
259 write_msa( out_msa, outfile )
260 write_msa( passed_seqs, passed_seqs_outfile )
261 write_msa( failed_seqs, failed_seqs_outfile )
264 write_msa( out_msa_pairs, outfile +"_" + min_linker.to_s )
268 write_msa( out_msa_singlets, outfile +"_singles" )
271 if out_msa_distant_partners
272 write_msa( out_msa_distant_partners, outfile + "_" + min_linker.to_s + "_isolated" );
277 log << "passing domains : " + domain_pass_counter.to_s + ld
279 log << "single domains : " + singlets_counter.to_s + ld
280 log << "domains in close pairs : " + close_pairs_counter.to_s + ld
281 log << "isolated domains : " + distant_pairs_counter.to_s + ld
283 log << "failing domains : " + domain_fail_counter.to_s + ld
284 log << "proteins with passing domains: " + proteins_with_passing_domains.to_s + ld
285 log << "proteins with failing domains: " + proteins_with_failing_domains.to_s + ld
288 return domain_pass_counter
295 def write_msa( msa, filename )
297 w = FastaWriter.new()
298 w.set_line_width( 60 )
301 io.write_to_file( msa, filename, w )
303 error_msg = "could not write to \"" + filename + "\""
304 raise IOError, error_msg
309 def add_sequence( sequence_name, in_msa, add_to_msa )
310 seqs = in_msa.find_by_name_start( sequence_name, true )
311 if ( seqs.length < 1 )
312 error_msg = "sequence \"" + sequence_name + "\" not found in sequence file"
313 raise StandardError, error_msg
315 if ( seqs.length > 1 )
316 error_msg = "sequence \"" + sequence_name + "\" not unique in sequence file"
317 raise StandardError, error_msg
319 seq = in_msa.get_sequence( seqs[ 0 ] )
320 add_to_msa.add_sequence( seq )
323 # raises ArgumentError, StandardError
324 def extract_domain( sequence,
333 add_domain_number_as_digit,
334 add_domain_number_as_letter,
337 if number.is_a?( Fixnum ) && ( number < 1 || out_of < 1 || number > out_of )
338 error_msg = "impossible: number=" + number.to_s + ", out of=" + out_of.to_s
339 raise ArgumentError, error_msg
341 if seq_from < 1 || seq_to < 1 || seq_from >= seq_to
342 error_msg = "impossible: seq-f=" + seq_from.to_s + ", seq-t=" + seq_to.to_s
343 raise ArgumentError, error_msg
345 seqs = in_msa.find_by_name_start( sequence, true )
347 error_msg = "sequence \"" + sequence + "\" not found in sequence file"
348 raise StandardError, error_msg
351 error_msg = "sequence \"" + sequence + "\" not unique in sequence file"
352 raise StandardError, error_msg
354 # hmmsearch is 1 based, wheres sequences are 0 bases in this package.
355 seq = in_msa.get_sequence( seqs[ 0 ] ).get_subsequence( seq_from - 1, seq_to - 1 )
357 orig_name = seq.get_name
359 seq.set_name( orig_name.split[ 0 ] )
362 seq.set_name( seq.get_name + "_" + seq_from.to_s + "-" + seq_to.to_s )
366 seq.set_name( seq.get_name[ 0, seq.get_name.length - TRIM_BY ] )
370 if add_domain_number_as_digit
371 seq.set_name( seq.get_name + number.to_s )
372 elsif add_domain_number_as_letter
374 error_msg = 'too many identical domains per sequence, cannot use letters to distinguish them'
375 raise StandardError, error_msg
377 seq.set_name( seq.get_name + ( number + 96 ).chr )
378 elsif add_domain_number
379 seq.set_name( seq.get_name + "~" + number.to_s + "-" + out_of.to_s )
383 # if ( seq.get_name.length > 10 )
384 # error_msg = "sequence name [" + seq.get_name + "] is longer than 10 characters"
385 # raise StandardError, error_msg
389 a = orig_name.rindex "["
390 b = orig_name.rindex "]"
392 error_msg = "species not found in " + orig_name
393 raise StandardError, error_msg
395 species = orig_name[ a .. b ]
396 seq.set_name( seq.get_name + " " + species )
398 out_msa.add_sequence( seq )
401 def is_ignorable?( line )
402 return ( line !~ /[A-Za-z0-9-]/ || line =~/^#/ )
405 end # class HmmscanDomainExtractor